PROJECT SUMMARY A major mechanism by which tumors survive from both chemotoxic and targeted therapies is hypothesized to be the result intra-tumoral heterogeneity. Tumor heterogeneity have been postulated to contribute to tumor evolution and therapy evasion through different mechanisms: generation of resistant subclones, cooperation between minor subclonal populations, minor subclones generating stromal signaling that favors tumor progression, etc. The genetic and epigenetic diversity that underlies tumor heterogeneity is siystematially missed by traditional genomic approaches. The overall goal of this project is to develop approaches to characterize functional dependencies and specific transcriptional programs in distinct tumor subpopulations, using single cell molecular profiling and analysis, with specific emphasis on clones and niches inducing relapse or progression. The first aim (1) of this project is to develop and implement new methods to capture tumor clonal representation and to integrate large-scale expression and mutational analysis in single cells. This will be done by conducting deep exome and RNA sequencing together with single-cell profiling to identify point mutations, indels, and gene fusions in tumor cells. The second aim of this project (2) is to identify epigenetically distinct cell niches with identical mutational spectra that either escape treatment or can regenerate the tumor. Clonally expanded tumor cells may be genetically identical with respect to the mutations identified in (1) but represent distinct epigenetic states, characterized by distinct gene expression signatures. The methods developed in (1) will be applied to residual tumor cells following therapy in PDX models to elucidate the drivers of drug resistance in epigenetically distinct cells. Finally the third aim (3) is to identify the molecular interactions that implement cell-cell communication processes, within a heterogenous tumor- microenviroment milieu. This will be achieved by developing novel information theoretic methodologies for the prioritization of specific ligands and receptors that mediate single cell communication and interaction. These methodologies will be specifically applied to elucidate novel tumor checkpoint inhibitors and other mechanisms that are relevant to immunotherapy applications.